Analysis of the Positively and Non‐Positively Selected Non‐Coding Sequences of Human Chromosome 16

Abstract

We propose the flanking sequences of the SNPs from the noncoding regions of chromosome 16 recently positively selected (Pos) exhibit regulatory processes different from sequences recently non‐positively selected (NonPos)(Nature 449.7164 (2007): 913–18). SNPs from the noncoding regions identified from UCSC Genome Browser Collection were screened using Haplotter in the Phase II HapMap Database found within all people groups. SNP flanking sequences from dbSNP build 131 with perfect matches were searched with the PATCH program of TRANSFAC database. Proximity Testing was done on these SNPs by analyzing their distances from nearest known genes. Of the 10537 and 18433 NonPos and Pos SNPs examined respectively, 74% and 76% of each are located at transcription factor related sites; 1.2% and 1.5% are unique. The proximity test rendered statistically significant data: NonPos SNPs are on average closer to genes than Pos SNPs at the same distance ranges. There are differences in the classes of regulatory functions between the two cohorts, (e.g. enhancers only in Unique Pos) bearing out our prediction. It seems reasonable to assume the Pos cohort should be closer to coding sequences. The result may prompt us to examine the appropriateness of using the chimp allele from pantro2 as ancestral allele in the selection algorithm. This work was supported by Wheaton College Alumni Association and Hong Kong Bioinformatics Center.